The school calls: Your child has an upset stomach, and you need to pick her up.
Did this come on suddenly? Is anyone else in her class sick? Was she playing normally at recess? What did the cafeteria serve for lunch?
As a parent, you want to know everything you can about the circumstances surrounding her illness.
Elisabeth Dowling Root shows the same diligence as a researcher studying children’s health at The Ohio State University’s Translational Data Analytics Institute.
“When we study children’s health, we need to incorporate and understand data,” she said, “not just where they were and what infectious agent they might have been exposed to, but what other children they came into contact with, and how their nutrition might also affect their immune function, or their age or how many children are in their household.
Children’s health cannot be understood without looking at it in a multifactorial, ecological way.
“Geography has a very unique view of the world. We see everything as interconnected. I’m connected to my environment and the people in my environment, and I’m also connected to policies and programs that I have no ability to impact, for the most part. Most medical studies are completely devoid of that information. A lot of my work has been convincing people to measure those variables.”
Part of Root’s work is based on quantifying socioeconomic factors that may play a role in determining how a child reacts to a medical intervention.
“That’s where big data comes in. Because now we’re not talking about one child and measuring how he or she is doing over time. My data sets are often only 20,000 kids, which is not big data in and of itself. But 20,000 kids measured for five years continuously, looking at 200, 300 or 400 variables in their lives — that blows it up big time.”
For example, several of Root’s projects have studied the effectiveness of vaccines in different countries around the world.
Using existing data on a medical intervention, Root links socioeconomic information about the children, their families and their environments — “complexities of the world that sort of mess up the way a vaccine trial is supposed to work in the field” — and creates statistical analyses that better examine how the vaccine is working.
Much of her research has found that after factoring in socioeconomic data, a vaccine actually works better than investigators initially thought.
“There are so many topics that we can use large data sets for that are out there waiting for us. We have to have researchers who understand how to use these data to examine these problems in depth with all the complications that surround them. This is what big data can do.”